Roundabout Capacity Calculator

This calculator uses the exponential-regression form of entry capacity that appears in the HCM's roundabout chapter: Entry Capacity = alpha times e to the power of negative beta times Conflicting Circulating Flow, where alpha and beta are empirically fit constants that switch based only on Entry Lanes and Circulating Lanes (this is one established HCM approach; some jurisdictions and later HCM editions instead use the more detailed NCHRP-derived linear regression form, which is not the model implemented here). Because that formula's only continuous input is Conflicting Circulating Flow, raising circulating traffic always lowers Entry Capacity. This calculator does not collect a separate entry-approach traffic count: Entry Volume, the demand side of the Volume-to-Capacity Ratio, is estimated internally as 60% of Conflicting Circulating Flow (a fixed placeholder ratio, not a measured field value), so V/C Ratio, Control Delay, Level of Service, and Queue Length are, in effect, all functions of Conflicting Circulating Flow alone -- rising circulating traffic simultaneously lowers Entry Capacity and raises the assumed Entry Volume, compounding to raise V/C Ratio faster than the capacity term alone would suggest. If your actual entry-approach demand differs meaningfully from 60% of the circulating flow you enter, treat V/C Ratio, Control Delay, Level of Service, and Queue Length as illustrative rather than site-specific, and verify against a real turning-movement count.

Entry Width, Inscribed Circle Diameter, and Entry Angle are real geometric design parameters that matter for roundabout performance in general, and they are collected here for context and cross-reference against the field measurements you took, but this specific exponential model does not feed them into the Entry Capacity, Delay, or Level of Service formulas at all -- only the lane-count-based alpha/beta pair and Conflicting Circulating Flow drive those outputs. If your design work depends on how apron width or a tighter inscribed diameter changes capacity, that requires the geometry-based HCM regression model instead of the exponential form used here. Control Delay adds a fixed five seconds of geometric delay on top of the queuing-theory result, and 95th-percentile Queue Length is a rough approximation, not a full HCM queuing-percentile calculation.